CORA: Adapting CLIP for Open-Vocabulary Detection with Region Prompting and Anchor Pre-Matching

Benchmark Model Rank Results
described-object-detection-on-description-detectionCORA-R50#8Intra-scenario FULL mAP: 6.2Intra-scenario PRES mAP: 6.7…
open-vocabulary-object-detection-on-mscocoCORA+#7AP 0.5: 43.1
open-vocabulary-object-detection-on-mscocoCORA#10AP 0.5: 41.7